.png){kind=icon}
Return
Toward Efficient Transparent Computing for IoT Apps by On-Chip Kernel Offload
.png){kind=icon}
PRE
AI
DOI：10.1109/JIOT.2018.2875050.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Graphics processing units (GPUs) have of late enjoyed increased popularity as a general purpose computing accelerator in multiple application domains like artificial intelligence. However, there has been little exploration into the performance and energy optimization GPUs can deliver for increasingly popular but heavy burden on transparent computing for Internet-of-Things (IoT), such as deep-inference tasks under energy-constrained scenarios. This paper presents FuncShare, an extension of the capability of dynamic functional connectivity among lot to cross-device hand over identified computation-intensive functional codes at runtime. As a kernel-level offloading solution, FuncShare now enables unmodified Android and Linux-based transparent computing applications (apps) to utilize not only application functionalities but also system functionalities across devices, as if they were to utilize them inside the same OpenCL or CUDA libraries. Toward secure network connection, FuncShare also allows performing of permission checks for remote apps in the same method as for local. Experimental results show that FuncShare enables transparent cross-device sharing for power-hungry functionalities and benefits significant performance enhancing than within-device processing even if a large amount of data is transferred.
Keywords:
Dynamic functional connectivity
graphics processing unit (GPU)
Internet-of-Things (IoT)
performance optimization
transparent computing
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
8.9
Papers:
1.4W
Citations:
7.8W
Organization
Cited Papers
ShareSave
Block-Stream as a Service: A More Secure, Nimble, and Dynamically Balanced Cloud Service Model for Ambient Computing.png)
IEEE NETWORK
IEEE NETWORKIF6.3
PREAI
ShareSaveShareSave
ShareSaveShareSave
View more